Use of transcriptomic and genomic data to predict risk of Abdominal Aortic Aneurysm

Use of transcriptomic and genomic data to predict risk of Abdominal Aortic Aneurysm

Lay summary

Background: Abdominal Aortic Aneurysms (AAA) consist on a local dilatation of the abdominal aorta, and are characterized by structural deterioration of the vascular wall leading to progressive dilatation and, potentially, rupture of the abdominal aorta. AAA are usually asymptomatic until they rupture, and no clinical treatment exist beyond surgery. Its prevalence ranges from 4-7% in men and 1-2% in women over 65. The risk of rupture is associated with aneurysm size and growth rate, and it results in death in about 90% of the ruptured cases, representing the 12-15th leading cause of dead in persons over 55 years of age in USA and Europe.

Genome-wide association studies have been used to identify new susceptibility genes for aortic aneurysm, which helped identify novel loci associated with AAA risk although novel findings explain a disappointingly small proportion of the risk of disease.

Aims: The present project aims to identify novel biological pathways underlying AAA by taking advantage of new methods to predict genetically regulated transcriptomes. Specifically, we propose to exploit transcriptomic and genomic data from several AAA human cohorts. We will then use novel methodologies that allow to better exploit global genomic and transcriptomic data in an integrated manner to understand disease.

Methods: In our project, we will generate models based on our available aortic tissue biobanks to get an accurate prediction of the expression in different cell types of the aorta based on genetic data. We will then test their performance in a large meta-analysis of AAA cases and controls (including UK-Biobank) to infer expression based on genetic information. Specifically, we will utilize an innovative bioinformatic tool (prediXcan) recently developed by the GTEx Consortium, that estimates the component of gene expression determined by an individual's genetic profile to identify genes involved in the etiology of disease. Effectively, this corresponds to calculating local polygenic scores for gene expression. The advantage is that information from transcriptome and genome is integrated, resulting in a marked gain in power.

Expected results: We expect that this strategy will result in a notable increase in statistical power and will, at the same time, inform about the biological pathways involved in disease. This is an unprecedented exercise that has never been done before in the area of aortic aneurysm, and which has the potential to reveal novel genes and biological pathways associated with the risk of AAA.